ABSTRACT
Background: Polycystic ovary syndrome (PCOS) is a complex syndrome with various metabolic disorders. Previous studies have reported that D-lactate is commonly related to metabolic disturbances. However, the role of D-lactate in PCOS remains unclear.
Methods: Serum D-lactate levels from control (n = 132) and women with PCOS (n = 132) were measured, and their relationship with several metabolic parameters were analyzed. The correlation of serum D-lactate with PCOS was assessed using logistic regression analysis, and the performance of serum D-lactate as a potential predictor for PCOS was evaluated using receiver operating characteristic curve analysis.
Results: Serum D-lactate levels were markedly higher in patients with PCOS compared with those of controls (P < 0.001). The proportion of PCOS was substantially higher in increasing quartiles of serum D-lactate levels (P < 0.001). After adjusting for other confounders, there was still a correlation between D-lactate and PCOS (P < 0.001; odds ratio, 5.654; 95% confidence interval, 3.091–10.342). D-lactate levels were positively correlated with fasting serum insulin, homeostasis model assessment of insulin resistance, and triglycerides in patients with PCOS (all P < 0.01), which was not detected in controls (P > 0.05). D-lactate had an area under the curve (AUC) of 79.4% in predicting PCOS, with a similar performance as anti-Müllerian hormone (AMH) and luteinizing hormone (LH), and its combination with AMH and LH yielded a higher AUC of 90.9%.
Conclusions: Substantially elevated serum D-lactate levels are significantly associated with PCOS, highlighting the importance of further research into the role of D-lactate in the pathogenesis of PCOS.
Key words: D-lactate, polycystic ovary syndrome, obesity, insulin resistance, dyslipidemia
INTRODUCTION
Polycystic ovary syndrome (PCOS) is a complex endocrine and metabolic disorder that affects approximately 5%–20% of women of childbearing age worldwide.[1–3] Monosaccharides are the simplest carbohydrates and principal energy sources of biological processes.[4] Using RNA sequencing technology, we have recently found that monosaccharide biosynthesis involved in the pathogenesis of PCOS.[5] Our previous study also showed that the serum levels of fructose, a major monosaccharide, were markedly elevated in women with PCOS.[6]
D-lactate, an optical isomer of lactic acid, is a downstream metabolic product of fructose.[7] Many studies have demonstrated that D-lactate is highly relevant to insulin resistance and obesity, which are common features of PCOS.[8–10] For instance, Erkin-Cakmak et al. showed that a decrease in D-lactate levels can improve insulin sensitivity in children with obesity and metabolic syndrome.[10] Elevated serum D-lactate levels are positively associated with insulin resistance and low-density lipoprotein cholesterol (LDL-C) levels in obese adolescents.[9] In addition, recent research suggests that urinary and blood levels of D-lactate are increased in patients with diabetes.[11] Considering the known role of insulin resistance and obesity in PCOS,[12–14] we hypothesized that D-lactate may be involved in the pathogenesis and progression of PCOS.
To date, no study has investigated the serum D-lactate levels of women with PCOS or the association between serum D-lactate and PCOS. This study aimed to measure and analyze the serum D-lactate levels in women with PCOS and healthy participants and elucidate the relationship between serum D-lactate and PCOS.
PATIENTS AND METHODS
Patients and samples collection
A total of 132 women with PCOS and 132 healthy women without PCOS (controls) were enrolled at the Shengjing Hospital of China Medical University from July 13 to December 20, 2020. Women with PCOS were diagnosed according to the new guidelines for the diagnosis and treatment of PCOS.[15] For the diagnosis of PCOS in adults, PCOS was present if at least two of the following three features were observed: biochemical or clinical evidence of hyperandrogenism, menstrual disturbance (oligomenorrhea, amenorrhea, or anovulation), and polycystic ovarian morphology on ultrasonography. The exclusion criteria are briefly described in our previous publication, which were the presence of endocrine diseases and recent hormonal medication.[16] The women without PCOS comprised the control group, and had clinical infertility due to fallopian tube or male factor infertility. Women with PCOS and the control group were matched for age and body mass index (BMI). Data on age and BMI were collected from the electronic medical record database of the Shengjing Hospital of China Medical University.
Venous blood samples were collected early in the morning under 12 h of fasting conditions on the third to fifth days of spontaneous menstruation or progestin-induced withdrawal bleeding. The blood was centrifuged at 2,000 × g for 10 minutes at 4°C, and the serum was divided into two parts, one for serum parameter detection in the clinical laboratory at Shengjing Hospital, which included fasting plasma glucose (FPG), fasting serum insulin (FSI), lipids, total testosterone (TT), follicle-stimulating hormone (FSH), estradiol, prolactin, progesterone, anti-Müllerian hormone (AMH), luteinizing hormone (LH), and thyroid-stimulating hormone (TSH), and the other for assessing free testosterone (FT), dehydroepiandrosterone sulfate (DHEAS), sex hormone-binding globulin (SHBG), and D-lactate in our internal laboratory. Repeated freezing and thawing were avoided throughout the experimental procedure.
The study was performed in accordance with the principles of the Declaration of Helsinki. The present study was approved by the Institutional Review Board of China Medical University (approval number 2020PS198K). All participants were exempted from providing informed consent, as the patient specimens used in the study were all discarded specimens after routine clinical diagnosis, and privacy-related information was removed.
Laboratory measurements
Serum levels of FPG, FSI, total cholesterol (TC), LDL-C, high-density lipoprotein cholesterol (HDL-C), and triglycerides (TG) were measured using enzymatic assays on an ARCHITECT ci16200 Automatic Biochemical Analyzer (Abbott Laboratories, USA) with corresponding kits (Architect glucose Reagent Kit, Architect insulin Reagent Kit, Abbott Laboratories, USA; and TC Test Kit, LDL-C Test Kit, HDL-C Test Kit, TG Test Kit, Kyowa Medex, Japan). The homeostasis model assessment of insulin resistance (HOMA-IR) index was calculated using the following formula: HOMA-IR = (FSI, μU/mL × FPG, mmol/L) / 22.5.
The concentrations of several hormones, including TT, FSH, estradiol, prolactin, progesterone, AMH, LH, and TSH, were assessed using chemiluminescence assays on an UniCel DxI 800 Automatic Immunoassay System (Beckman Coulter, USA) with commercial kits (C10160, C10155, C10156, C10161, C10159, C10158, B13127, C10153, Beckman Coulter, USA).
DHEAS (Cusabio Biotech, Wuhan, China; CSB-E05 105 h), FT (Cusabio Biotech; CSB-E05 096 h), and SHBG (Human SHBG ELISA Kit; RayBiotech, Norcross, GA, USA) were measured using commercially available enzyme-linked immunosorbent assay kits according to the manufacturer's instructions. The assay's sensitivity limits for detecting DHEAS, FT, and SHBG were defined by the manufacturer as 10 μg/L, 3.75 ng/L, and 1.2 pmol/L, respectively. The intra-assay coefficients of variation (CVs) were 6.8% for FT, 5.5% for DHEAS, and 9.2% for SHBG, and the inter-assay CVs were 10.2%, 8.3%, and 6.1%, respectively. Serum concentrations were calculated by comparing the optical density (450 nm) of each sample with that of the standard curve.
D-lactate concentration was measured using a D-lactate fluorometric assay kit (BioVision Inc., Milpitas, CA, USA; K668–100). The assay was performed as per the manufacturer's instructions. By recognizing a fluorescent signal (Ex/Em = 535/587 nm), the serum concentration of D-lactate can be computed using the formula listed in the instructions. The intra-and inter-assay CVs were 5.8% and 8.7%, respectively.
Participants grouping
To explore whether the difference of D-lactate levels between women with PCOS and control women was attributed to several D-lactate-related metabolic disorders, including obesity, insulin resistance, and dyslipidemia, the participants were further classified into subgroups.
According to the criteria proposed by the World Health Organization, overweight and obesity in Asian women were defined as BMIs between 23 and 25, and over 25, respectively.[17] To explore whether the difference in D-lactate levels between PCOS and control groups was attributed to obesity, PCOS and control groups were subdivided into lean, overweight, and obese subgroups (BMI < 23, 23–25, and > 25;
Insulin resistance was defined as an HOMA-IR of more than 2.5.[18] To investigate whether the difference in D-lactate levels between PCOS and control groups was due to insulin resistance, PCOS and control groups was sub-classified into the non-insulin-resistant and insulin-resistant subgroups (HOMA-IR < 2.5, and ≥ 2.5;
Dyslipidemia was defined as TC ≥ 6.2 mmol/L, TG ≥ 2.3 mmol/L, LDL-C ≥ 4.1 mmol/L, HDL-C < 1.0 mmol/L, fulfilling at least one criteria described above.[19] To investigate whether the difference in D-lactate levels between PCOS and control groups was due to the presence of dyslipidemia, these two groups were subdivided into dyslipidemia and normolipidemia groups (
Statistical analysis
Statistical Package for Social Sciences, version 22.0 (SPSS 22.0) (IBM Corporation, Armonk, NY, USA) was used to analyze the acquired values. The normal distribution of continuous data was examined using the Kolmogorov–Smirnov test. Normal distribution variables were presented as mean ± standard deviation and were compared using Z-test. Non-normal distribution variables were presented as the median (interquartile range) and were compared using the Mann–Whitney U-test. The total participants were grouped based on quartile D-lactate levels, and differences among these four groups were analyzed with one-way analysis of variance or the Kruskal–Wallis test for continuous variables. Categorical variables were compared using the Chi–square test.
Univariate logistic regression analysis was used to investigate the covariates affecting PCOS, and multivariate logistic regression analysis was carried out on the significant variables defined by univariate analysis. The correlations of D-lactate concentration with all the other clinical indicators were evaluated with Pearson's or Spearman correlation coefficient analyses.
The performance of D-lactate, AMH, LH, and other parameters for predicting PCOS was evaluated by analyzing the receiver operating characteristic (ROC) curves. The area under the curve (AUC) was used to calculate the sensitivity, specificity, and 95% confidence interval (CI) for the prediction of PCOS. The performance of D-lactate, AMH, LH, and the combination of D-lactate with AMH or LH in predicting PCOS was compared according to the method proposed by Hanley and McNeil.[20] All statistical tests were two-sided, and statistical significance was set at P < 0.05 between the two groups.
RESULTS
Women with PCOS presented with higher serum D-lactate levels than control participants
Serum D-lactate levels in women with PCOS were markedly higher than those in control participants (3.59 ± 0.77 pmol/µL vs. 2.73 ± 0.71 pmol/µL; P < 0.001; Table 1; Figure 1A). To explore whether the difference in D-lactate levels between PCOS and control groups was attributed to several D-lactate-related metabolic disorders, serum D-lactate levels between PCOS and control groups were further compared in several metabolic subgroups.
Figure 1. Serum D-lactate levels in women with PCOS and control participants. (A) Differences in serum D-lactate levels between women with PCOS and control participants. (B) Differences in serum D-lactate levels between women with PCOS and the control participants in the lean, overweight and obese subgroups. (C) Differences in serum D-lactate levels between women with PCOS and the control participants in the insulin-resistant and noninsulin-resistant subgroups. (D) Differences in serum D-lactate levels between women with PCOS and the control participants in the dyslipidemia and normolipidemia subgroups. **P < 0.01. PCOS, polycystic ovary syndrome.
| Clinical data | Control group (N = 132) | PCOS group (N = 132) | P-value |
| Age (years) | 31.00 (29.00-33.75) | 31.00 (28.00-33.00) | 0.354 |
| BMI | 24.60 (22.05-27.48) | 24.50 (21.90-26.95) | 0.618 |
| D-lactate (pmol/µL) | 2.73 ± 0.71 | 3.59 ± 0.77 | < 0.001 |
| TT (µg/L) | 0.48 (0.34-0.64) | 0.63 (0.48-0.79) | < 0.001 |
| FT (nmol/L) | 0.03 (0.02-0.04) | 0.04 (0.02-0.05) | < 0.001 |
| SHBG (nmol/L) | 31.40 (22.25-51.65) | 24.90 (17.35-36.85) | 0.001 |
| DHEAS (µmol/L) | 3.18 (2.17-4.46) | 4.65 (2.87-7.00) | < 0.001 |
| AMH (pmol/L) | 27.99 (13.85-39.84) | 62.05 (39.70-87.82) | < 0.001 |
| FSH (mIU/mL) | 7.24 ± 2.25 | 6.56 ± 1.95 | 0.010 |
| LH (mIU/mL) | 3.83 (3.01-5.62) | 10.22 (5.74-15.49) | < 0.001 |
| Estradiol (ng/L) | 43.00 (33.00-56.00) | 47.00 (36.00-68.00) | 0.086 |
| Prolactin (μg/L) | 11.85 (8.60-16.64) | 10.02 (7.58-13.77) | 0.004 |
| P (µg/L) | 0.55 (0.40-0.81) | 0.58 (0.40-1.05) | 0.281 |
| TSH (µIU/mL) | 1.84 (1.29-2.69) | 1.89 (1.31-2.62) | 0.924 |
| FPG (mmol/L) | 5.31 (4.91-5.66) | 5.18 (4.92-5.57) | 0.241 |
| FSI (mIU/L) | 10.55 (8.43-14.50) | 12.40 (8.90-18.90) | 0.055 |
| HOMA-IR | 2.53 (1.93-3.56) | 2.94 (1.94-4.37) | 0.113 |
| TC (mmol/L) | 4.65 (4.05-5.20) | 4.63 (4.21-5.34) | 0.840 |
| LDL-C (mmol/L) | 2.93 ± 0.85 | 2.89 ± 0.76 | 0.804 |
| HDL-C (mmol/L) | 1.21 (1.05-1.41) | 1.19 (1.04-1.35) | 0.346 |
| TG (mmol/L) | 0.97 (0.69-1.56) | 1.27 (0.87-2.10) | 0.003 |
In the lean, overweight, and obese subgroups, circulating D-lactate levels were all higher in women with PCOS than those in the control participants (P < 0.001, P= 0.002, P < 0.001; Figure 1B;
Higher serum D-lactate levels correlated with PCOS
We focused on the correlation between serum D-lactate levels and PCOS. All participants were grouped according to quartiles of serum D-lactate levels (Table 2). The proportion of PCOS was substantially higher combined with increasing quartiles of serum D-lactate levels (P < 0.001). After adjusting for age, BMI, and the covariates affecting PCOS identified in the univariate logistic regression analysis, including TT, SHBG, AMH, FSH, LH, FSI, HOMA-IR, and TG, there was still a significant correlation between the circulating D-lactate levels and PCOS (P < 0.001; odds ratio (OR) 5.654; 95% CI, 3.091–10.342; Table 3).
| Characteristics | Quartile 1 (N = 66) | Quartile 2 (N = 68) | Quartile 3 (N = 67) | Quartile 4 (N = 63) | P-value |
| D-lactate (pmol/µL) | ≤ 2.57 | > 2.57 and ≤ 3.13 | > 3.13 and ≤ 3.84 | > 3.84 | |
| PCOS | 10 (15.2%) | 25 (36.8%) | 43 (64.2%) | 54 (85.7%) | < 0.001 |
| Age (years) | 31.00 (29.00-33.00) | 31.00 (28.00-33.75) | 30.00 (28.00-33.00) | 32.00 (29.00-34.00) | 0.213 |
| BMI | 22.95 (21.68-25.68) | 23.35 (21.46-26.08) | 25.40 (22.40-28.00) | 25.80 (23.50-28.00) | < 0.001 |
| TT (μg/L) | 0.49 (0.35-0.66) | 0.50 (0.38-0.68) | 0.59 (0.46-0.84) | 0.63 (0.47-0.76) | 0.005 |
| FT (nmol/L) | 0.03 (0.02-0.04) | 0.03 (0.02-0.04) | 0.03 (0.02-0.05) | 0.03 (0.02-0.05) | 0.032 |
| SHBG (nmol/L) | 35.05 (23.33-57.05) | 31.50 (23.10-50.90) | 24.75 (18.75-32.35) | 24.10 (16.33-40.68) | 0.001 |
| DHEAS (µmol/L) | 3.31 (2.20-4.80) | 3.38 (2.26-5.71) | 4.24 (2.85-6.02) | 4.18 (2.83-6.50) | 0.015 |
| AMH (pmol/L) | 29.77 (17.49-43.77) | 39.56 (27.77-68.47) | 39.84 (20.71-69.04) | 56.76 (32.06-87.11) | < 0.001 |
| FSH (mIU/mL) | 7.43 ± 2.25 | 6.91 ± 2.14 | 6.83 ± 2.08 | 6.39 ± 1.95 | 0.049 |
| LH (mIU/mL) | 4.26 (3.01-7.32) | 4.79 (3.46-12.14) | 7.36 (3.61-12.62) | 8.99 (4.65-12.17) | 0.002 |
| Estradiol (ng/L) | 43.00 (34.00-59.00) | 45.00 (34.00-67.00) | 43.00 (32.00-59.75) | 46.00 (34.00-57.00) | 0.987 |
| Prolactin (μg/L) | 12.66 (9.04-17.33) | 10.61 (7.66-14.10) | 9.74 (7.72-14.88) | 10.21 (7.48-14.55) | 0.014 |
| P (μg/L) | 0.54 (0.39-0.97) | 0.59 (0.45-0.82) | 0.57 (0.40-1.03) | 0.54 (0.37-1.07) | 0.892 |
| TSH (µIU/mL) | 2.08 (1.35-2.79) | 1.86 (1.20-2.75) | 1.53 (1.07-2.66) | 1.93 (1.41-2.56) | 0.189 |
| FPG (mmol/L) | 5.27 (4.91-5.67) | 5.16 (4.84-5.59) | 5.21 (4.96-5.63) | 5.20 (4.93-5.52) | 0.734 |
| FSI (mIU/L) | 9.70 (8.23-12.43) | 10.80 (8.13-14.30) | 13.60 (9.30-19.60) | 14.00 (10.10-18.90) | 0.001 |
| HOMA-IR | 2.31 (1.85-2.96) | 2.38 (1.81-3.42) | 3.16 (2.08-4.63) | 3.05 (2.35-4.27) | 0.002 |
| TC (mmol/L) | 4.54 (4.03-5.04) | 4.53 (3.98-5.24) | 4.61 (4.23-5.25) | 5.00 (4.40-5.50) | 0.056 |
| LDL-C (mmol/L) | 2.82 ± 0.81 | 2.86 ± 0.92 | 2.90 ± 0.68 | 3.06 ± 0.78 | 0.327 |
| HDL-C (mmol/L) | 1.22 (1.05-1.44) | 1.21 (1.00-1.39) | 1.21 (1.07-1.36) | 1.16 (1.04-1.32) | 0.743 |
| TG (mmol/L) | 0.96 (0.65-1.48) | 1.21 (0.77-1.74) | 1.12 (0.72-1.81) | 1.48 (0.98-2.51) | 0.001 |
| Characteristics | Univariate regression | Multivariate regression | ||||
| OR | 95%CI | P-value | OR | 95%CI | P-value | |
| Age (years) | 0.980 | 0.922-1.042 | 0.514 | 1.056 | 0.953-1.170 | 0.298 |
| BMI | 0.978 | 0.914-1.047 | 0.523 | 0.833 | 0.735-0.944 | 0.004 |
| TT (µg/L) | 12.721 | 4.049-39.964 | < 0.001 | 2.494 | 0.467-13.306 | 0.285 |
| AMH (pmol/L) | 1.391 | 1.272-1.522 | < 0.001 | 1.259 | 1.119-1.416 | < 0.001 |
| FSH (mIU/mL) | 0.855 | 0.757-0.965 | 0.011 | 0.813 | 0.662-0.998 | 0.047 |
| LH (mIU/mL) | 1.277 | 1.191-1.371 | < 0.001 | 1.192 | 1.096-1.297 | < 0.001 |
| Prolactin (μg/L) | 0.934 | 0.891-0.978 | 0.004 | 1.000 | 0.934-1.070 | 0.995 |
| FSI (mIU/L) | 1.038 | 1.001-1.076 | 0.041 | 1.044 | 0.983-1.109 | 0.161 |
| TG (mmol/L) | 1.262 | 1.038-1.533 | 0.019 | 1.245 | 0.952-1.629 | 0.110 |
| D-lactate (pmol/µL) | 4.744 | 3.127-7.197 | < 0.001 | 5.654 | 3.091-10.342 | < 0.001 |
The correlation of D-lactate with the hormonal and metabolic parameters in PCOS and non-PCOS groups were further detected, respectively (Table 4). Spearman correlation coefficient analyses revealed that D-lactate levels of women with PCOS were positively correlated with BMI, FSI, HOMA-IR, and TG (all P < 0.01). Among these indicators, BMI was the most related parameter with D-lactate in women with PCOS (R = 0.290, P < 0.001). However, there was no correlation between D-lactate and FSI, HOMA-IR, and TG in the non-PCOS group (all P > 0.05). Collectively, higher serum D-lactate levels correlated with PCOS, and was associated with FSI, HOMA-IR, and TG only in women with PCOS.
| Characteristics | PCOS (N = 132) | Control (N = 132) | |||
| R | P-value | R | P-value | ||
| Age (years) | 0.095 | 0.119 | 0.012 | 0.845 | |
| BMI | 0.290 | < 0.001 | 0.170 | 0.004 | |
| TT (μg/L) | 0.022 | 0.717 | 0.070 | 0.237 | |
| FT (nmol/L) | 0.039 | 0.556 | 0.057 | 0.390 | |
| SHBG (nmol/L) | -0.105 | 0.106 | -0.089 | 0.143 | |
| DHEAS (µmol/L) | -0.031 | 0.606 | 0.028 | 0.642 | |
| AMH (pmol/L) | 0.048 | 0.419 | -0.061 | 0.303 | |
| FSH (mIU/mL) | -0.064 | 0.283 | -0.034 | 0.565 | |
| LH (mIU/mL) | -0.072 | 0.226 | 0.009 | 0.880 | |
| Estradiol (ng/L) | -0.048 | 0.418 | -0.008 | 0.889 | |
| Prolactin (μg/L) | -0.037 | 0.528 | -0.114 | 0.057 | |
| P (μg/L) | -0.049 | 0.410 | -0.111 | 0.065 | |
| TSH (µIU/mL) | 0.045 | 0.446 | -0.033 | 0.571 | |
| FPG (mmol/L) | 0.098 | 0.101 | -0.028 | 0.631 | |
| FSI (mIU/L) | 0.210 | < 0.001 | 0.081 | 0.171 | |
| HOMA-IR | 0.211 | < 0.001 | 0.068 | 0.252 | |
| TC (mmol/L) | 0.111 | 0.061 | 0.120 | 0.043 | |
| LDL-C (mmol/L) | 0.110 | 0.064 | 0.110 | 0.063 | |
| HDL-C (mmol/L) | -0.084 | 0.161 | 0.012 | 0.846 | |
| TG (mmol/L) | 0.155 | 0.009 | 0.075 | 0.204 | |
The performance of serum D-lactate levels in predicting PCOS
Considering the higher serum D-lactate levels in women with PCOS and the correlation between D-lactate and PCOS, we explored the clinical value of serum D-lactate levels for PCOS. We first compared the performance of all the parameters included in this study for predicting PCOS using the ROC curves (Table 5). AMH, LH, and D-lactate presented the highest area under the curve (AUCs) in predicting PCOS among all parameters, which had significantly higher AUCs than testosterone, the typical markers for PCOS. We then compared the performance of D-lactate with AMH and LH in predicting PCOS, and explored the performance of their combined use (Figure 2).
Figure 2. Predictive performance of serum D-lactate levels in women with PCOS. (A) ROC curves illustrate the value of serum D-lactate, AMH, LH levels in predicting PCOS. (B) ROC curves for D-lactate, AMH, and a combination of AMH and D-lactate levels in predicting PCOS. (C) ROC curves for D-lactate, LH, and a combination of LH and D-lactate levels in predicting PCOS. (D) ROC curves for D-lactate, AMH, LH, and a combination of AMH, LH and D-lactate levels in predicting PCOS. (E) AUC, cutoff value, SE, 95% CI, sensitivity, specificity, PPV, and NPV for each ROC curve in this evaluation. PCOS, polycystic ovary syndrome; ROC, receiver operating characteristic; AMH, anti-Müllerian hormone; LH, luteinizing hormone; AUC, area under the curve; SE, standard error; CI, confidence interval; PPV, positive predictive value; NPV, negative predictive value.
| Characteristics | AUC | SE | Cut off | 95% CI | Sensitivity (%) | Specificity (%) | P-value |
| Age (years) | 0.467 | 0.036 | 25.50 | 0.397-0.537 | 94.7 | 10.6 | 0.356 |
| BMI | 0.482 | 0.036 | 23.05 | 0.412-0.552 | 65.2 | 38.6 | 0.618 |
| D-lactate (pmol/µL) | 0.794* | 0.028 | 3.14 | 0.739-0.848 | 73.1 | 75.6 | < 0.001 |
| TT (μg/L) | 0.672 | 0.033 | 0.48 | 0.607-0.737 | 77.9 | 49.2 | < 0.001 |
| FT (nmol/L) | 0.635 | 0.035 | 0.04 | 0.567-0.704 | 52.8 | 72.2 | < 0.001 |
| SHBG (nmol/L) | 0.379 | 0.037 | 95.40 | 0.307-0.452 | 6.4 | 94.4 | 0.001 |
| DHEAS (µmol/L) | 0.684 | 0.034 | 10.15 | 0.617-0.750 | 64.8 | 68.0 | < 0.001 |
| AMH (pmol/L) | 0.817* | 0.026 | 5.95 | 0.766-0.868 | 72.3 | 78.0 | < 0.001 |
| FSH (mIU/mL) | 0.418 | 0.035 | 1.90 | 0.349-0.488 | 99.2 | 0.8 | 0.023 |
| LH (mIU/mL) | 0.804* | 0.029 | 7.25 | 0.748-0.860 | 70.0 | 83.5 | < 0.001 |
| Estradiol (ng/L) | 0.562 | 0.036 | 53.5 | 0.492-0.632 | 38.9 | 74.2 | 0.086 |
| Prolactin (μg/L) | 0.397 | 0.035 | 0.37 | 0.328-0.465 | 100.0 | 0 | 0.004 |
| P (μg/L) | 0.539 | 0.036 | 0.98 | 0.468-0.609 | 29.0 | 83.6 | 0.281 |
| TSH (µIU/mL) | 0.497 | 0.036 | 1.43 | 0.427-0.567 | 70.5 | 36.4 | 0.924 |
| FPG (mmol/L) | 0.458 | 0.036 | 6.32 | 0.388-0.528 | 6.9 | 96.2 | 0.241 |
| FSI (mIU/L) | 0.568 | 0.035 | 14.75 | 0.499-0.638 | 40.2 | 77.3 | 0.055 |
| HOMA-IR | 0.557 | 0.036 | 2.89 | 0.487-0.627 | 53.1 | 60.6 | 0.113 |
| TC (mmol/L) | 0.507 | 0.036 | 4.07 | 0.437-0.577 | 80.8 | 25.8 | 0.840 |
| LDL-C (mmol/L) | 0.498 | 0.036 | 2.39 | 0.428-0.569 | 77.7 | 29.5 | 0.963 |
| HDL-C (mmol/L) | 0.466 | 0.036 | 1.11 | 0.396-0.536 | 66.2 | 36.4 | 0.346 |
| TG (mmol/L) | 0.607 | 0.035 | 0.98 | 0.539-0.675 | 68.5 | 50.8 | 0.003 |
The AUC of D-lactate for PCOS prediction was 79.4%, with a sensitivity of 73.1%, and a specificity of 75.2%, and the cutoff value was set at 3.14 pmol/µL (Figure 2A, E). As for the typical markers AMH and LH in predicting PCOS, AMH had an AUC of 81.7%, a sensitivity of 72.3%, and a specificity of 78.0% (Figure 2A, E). LH could predict PCOS with an AUC of 80.4%, a sensitivity of 70.0%, and a specificity of 83.5% (Figure 2A, E). When we compared the performance of these parameters with the method proposed by Hanley and McNeil,[20] there was no difference in terms of performance between D-lactate, AMH, and LH predicting PCOS (
The performance of combining serum D-lactate and AMH or LH in predicting PCOS was further explored. The combination of D-lactate and AMH produced an AUC of 88.4%, a sensitivity of 80.0%, and a specificity of 78.7% in PCOS prediction (Figure 2B, E). The AUC of combining D-lactate and LH was 88.3%, with a sensitivity of 79.2% and a specificity of 83.5% in predicting PCOS (Figure 2C, E). The combination of D-lactate, AMH, and LH yielded the highest AUC of 90.9%, with a sensitivity of 74.6% and a specificity of 91.3% for predicting PCOS (Figure 2D, E). The combination of D-lactate with AMH, LH, or both AMH and LH had a better predictive performance than AMH or LH alone (
As described previously, women with PCOS all presented with higher serum D-lactate levels regardless of their metabolic status. We were interested in exploring whether D-lactate and the combination model of D-lactate, AMH, and LH had certain suitability for people with different metabolic statuses. Therefore, we compared their performance in predicting PCOS in several metabolic subgroups (Table 6). AUC values varied depending on metabolic phenotypes but remained in the moderately to highly accurate range for all populations. Especially in obese, insulin-resistant and dyslipidemia population, D-lactate produced an AUC of 86.2%, 83.4%, and 80.2% in predicting PCOS, respectively. Moreover, the combination of D-lactate, AMH, and LH yielded an AUC of over 90.0% in the obese, lean, insulin-resistant, non-insulin-resistant and normolipidemia subgroups. Taken together, these data suggest that the serum D-lactate levels had a similar predictive performance to AMH and LH in PCOS identification, and its combined use with AMH and LH improved the performance of AMH and LH. Moreover, D-lactate and its combination model with AMH and LH were likely applicable across populations with different metabolic statuses.
| Metabolic statuses | D-lactate | AMH | LH | D-lactate+AMH | D-lactate+LH | D-lactate+AMH+LH |
| Obese | 0.862 (0.786-0.919) | 0.788 (0.702-0.858) | 0.779 (0.693-0.851) | 0.903 (0.835-0.950) | 0.901 (0.832-0.948) | 0.919 (0.854-0.961) |
| Overweight | 0.738 (0.592-0.853) | 0.846 (0.714-0.933) | 0.766 (0.623-0.875) | 0.867 (0.739-0.947) | 0.845 (0.713-0.932) | 0.893 (0.772-0.963) |
| Lean | 0.767 (0.666-0.849) | 0.850 (0.760-0.916) | 0.851 (0.761-0.917) | 0.870 (0.783-0.931) | 0.888 (0.805-0.944) | 0.910 (0.832-0.960) |
| IR | 0.834 (0.762-0.891) | 0.812 (0.737-0.872) | 0.787 (0.711-0.852) | 0.896 (0.834-0.941) | 0.886 (0.822-0.933) | 0.910 (0.851-0.952) |
| Non-IR | 0.720 (0.628-0.801) | 0.853 (0.774-0.912) | 0.842 (0.762-0.904) | 0.864 (0.787-0.921) | 0.873 (0.798-0.928) | 0.903 (0.833-0.951) |
| Dyslipidemia | 0.802 (0.705-0.879) | 0.702 (0.596-0.794) | 0.775 (0.674-0.856) | 0.812 (0.715-0.887) | 0.870 (0.782-0.932) | 0.851 (0.759-0.917) |
| Normolipidemia | 0.784 (0.714-0.844) | 0.872 (0.811-0.918) | 0.828 (0.762-0.882) | 0.918 (0.865-0.955) | 0.890 (0.832-0.933) | 0.935 (0.886-0.967) |
DISCUSSION
In the current study, we reported for the first time that serum D-lactate levels in women with PCOS were distinctly higher than those in the control group and were associated with PCOS even after adjusting for covariates affecting PCOS. This prompted us to further investigate the potential role of serum D-lactate levels in women with PCOS.
Our previous study found that women with PCOS had abnormal monosaccharide metabolism, and their serum fructose levels were significantly higher than those of healthy participants.[6] Fructose upregulates aldolase B to increase methylglyoxal production in the liver, and excess methylglyoxal can be converted by glyoxalase to D-lactate.[21] Consistent with this finding, short-term restriction of dietary fructose intake in obese children can result in reduced levels of D-lactate.[10] Therefore, we speculate that active monosaccharide metabolism in PCOS may promote more conversion of fructose to D-lactate, resulting in high D-lactate levels in women with PCOS.
Recent studies have confirmed that the imbalance of the gut microbiota correlate closely with the initiation of PCOS and its metabolic disorders.[22–24]Bacteroides and Lactobacillus are both common bacterial alterations in women with PCOS and were proved to dominate the production of D-lactate in the gut.[25–27] Thus, another contributor to the higher D-lactate levels in women with PCOS may be the increased production of D-lactate metabolites in the gastrointestinal tract due to the imbalance in the gut microbiota. Moreover, as the product of gut bacterial, D-lactate could induce the inflammatory process and facilitate the systemic inflammatory.[27–29] Interestingly, chronic inflammation is considered a key contributor to the pathogenesis of PCOS.[22,30] Although the precise mechanisms of the relationship of D-lactate with gut microbiota disturbances and chronic inflammation in PCOS remain unclear, serum D-lactate levels might reflect gut microbiota disturbances as well as the chronic inflammation status in women with PCOS to some extent.
An interesting finding of this study is that the circulating D-lactate levels were positively correlated with FSI, HOMA-IR, and TG in PCOS, and these interactions seemed unique to women with PCOS. Previous studies provided evidence to support our results.[9,31] In obese adolescents, higher serum D-lactate levels are strongly associated with serum triglyceride and low-density lipoprotein cholesterol levels.[9] D-lactate can promote adipocyte differentiation and contribute to obesity.[31] Obesity can aggravate PCOS symptoms, especially exacerbating insulin resistance.[32–34] Collectively, the potential role of D-lactate in PCOS may be its effects on the metabolic disorders of PCOS, which need to be clarified in follow-up research.
Although studies have reported that the serum levels of lactate, an optical isomer of D-lactate, were elevated in women with PCOS,[35,36] so far no study has investigated the circulating D-lactate levels in women with PCOS. The major strength of this study is that this is the first study to assess the serum D-lactate levels in women with PCOS and evaluate the relationship between serum D-lactate and PCOS. PCOS is a heterogeneous disorder with diverse clinical metabolic phenotypes.[12,13] Another strength of this study is that this study systematically analyzed the D-lactate levels among people with different metabolic statuses. We further confirmed that serum D-lactate levels were likely applicable for predicting PCOS across populations with different metabolic statuses. Additionally, our study has limitations that should be considered. The sample size and non-longitudinal design of the present study were limited; therefore, large-scale, longitudinal, and multi-central trials are needed to validate these findings.
CONCLUSION
In conclusion, substantially elevated serum D-lactate levels are significantly associated with PCOS, which provide new insights for further research on the biological role of D-lactate in PCOS.
Declaration
Acknowledgement
The authors thank the staff of the Center of Reproductive Medicine in Shengjing Hospital of China Medical University for their cooperation and support.
Author contributions
Yang HY, Song JH, Feng D and Shi B conceived and designed the study. Yang HY and Shi B contributed to the sample handling, data management, and analysis. Yang HY, Song JH, and Feng D wrote the paper. Yang HY, Song JH, Feng D and Shi B have accessed and verified the data. All authors had full access to all the data in the study and had final responsibility for the decision to submit for publication.
Source of funding
None.
Ethical approval
The study was performed in accordance with the principles of the Declaration of Helsinki. The present study was approved by the Institutional Review Board of China Medical University (approval number 2020PS198K).
Informed consent
All participants were exempted from providing informed consent, as the patient specimens used in the study were all discarded specimens after routine clinical diagnosis, and privacy-related information was removed.
Conflict of interest
The authors declare no competing interest.
Use of large language models, AI and machine learning tools
AI tools were not used in this paper.
Data availability statement
The analyzed data sets generated during the present study are available from the corresponding author on reasonable request.
REFERENCES
- Azziz R, Carmina E, Chen Z, et al. Polycystic ovary syndrome. Nat Rev Dis Primers. 2016;2:16057. DOI: 10.1038/nrdp.2016.57
- Helvaci N, Yildiz BO. Polycystic ovary syndrome as a metabolic disease. Nat Rev Endocrinol. 2025;21(4):230–244. DOI: 10.1038/s41574-024-01057-w
- Walter K. What Is Polycystic Ovary Syndrome? JAMA. 2022;327(3):294. DOI: 10.1001/jama.2021.19776
- Chandel NS. Carbohydrate Metabolism. Cold Spring Harb Perspect Biol. 2021;13(1):a040568. DOI: 10.1101/cshperspect.a040568
- Jiao J, Shi B, Wang T, et al. Characterization of long non-coding RNA and messenger RNA profiles in follicular fluid from mature and immature ovarian follicles of healthy women and women with polycystic ovary syndrome. Hum Reprod. 2018;33(9):1735–1748. DOI: 10.1093/humrep/dey255
- Shi B, Feng D, Sagnelli M, et al. Fructose levels are elevated in women with polycystic ovary syndrome with obesity and hyperinsulinemia. Hum Reprod. 2020;35(1):187–194. DOI: 10.1093/humrep/dez239
- Masterjohn C, Park Y, Lee J, Noh SK, Koo SI, Bruno RS. Dietary fructose feeding increases adipose methylglyoxal accumulation in rats in association with low expression and activity of glyoxalase-2. Nutrients. 2013;5(8):3311–3328. DOI: 10.3390/nu5083311
- Fang H, Anhê FF, Zada DK, et al. Gut substrate trap of D-lactate from microbiota improves blood glucose and fatty liver disease in obese mice. Cell Metab. 2025;37(9):1806-1819.e7. DOI: 10.1016/j.cmet.2025.07.001
- Rodríguez-Mortera R, Luevano-Contreras C, Solorio-Meza S, et al. Higher D-lactate levels are associated with higher prevalence of small dense low-density lipoprotein in obese adolescents. Clin Chem Lab Med. 2018;56(7):1100-1108. DOI: 10.1515/cclm-2017-0733
- Erkin-Cakmak A, Bains Y, Caccavello R, et al. Isocaloric Fructose Restriction Reduces Serum d-Lactate Concentration in Children With Obesity and Metabolic Syndrome. J Clin Endocrinol Metab. 2019;104(7):3003-3011. DOI: 10.1210/jc.2018-02772
- Maasen K, Hanssen NMJ, van der Kallen CJH, Stehouwer CDA, van Greevenbroek MMJ, Schalkwijk CG. Polymorphisms in Glyoxalase I Gene Are Not Associated with Glyoxalase I Expression in Whole Blood or Markers of Methylglyoxal Stress: The CODAM Study. Antioxidants (Basel). 2021;10(2):219. DOI: 10.3390/antiox10020219
- Amato MC, Vesco R, Vigneri E, Ciresi A, Giordano C. Hyperinsulinism and polycystic ovary syndrome (PCOS): role of insulin clearance. J Endocrinol Invest. 2015;38(12):1319–1326. DOI: 10.1007/s40618–015–0372-x
- Kakoly NS, Khomami MB, Joham AE, et al. Ethnicity, obesity and the prevalence of impaired glucose tolerance and type 2 diabetes in PCOS: a systematic review and meta-regression. Hum Reprod Update. 2018;24(4):455–467. DOI: 10.1093/humupd/dmy007
- Zhao H, Zhang J, Cheng X, Nie X, He B. Insulin resistance in polycystic ovary syndrome across various tissues: an updated review of pathogenesis, evaluation, and treatment. J Ovarian Res. 2023;16(1):9. DOI: 10.1186/s13048–022–01091–0
- Teede HJ, Tay CT, Laven J, et al. Recommendations from the 2023 International Evidence-based Guideline for the Assessment and Management of Polycystic Ovary Syndrome. Fertil Steril. 2023;120(4):767–793. DOI: 10.1016/j.fertnstert.2023.07.025
- Feng D, Shi B, Bi F, et al. Elevated Serum Mannose Levels as a Marker of Polycystic Ovary Syndrome. Front Endocrinol (Lausanne). 2019;10:711. DOI: 10.3389/fendo.2019.00711
- Dokras A, Luque-Ramírez M, Escobar-Morreale HF. POLYCYSTIC OVARY SYNDROME: ORIGINS AND IMPLICATIONS: Long-term health outcomes in polycystic ovary syndrome. Reproduction. 2025;170(2):e250118. DOI: 10.1530/REP-25–0118
- Parcha V, Heindl B, Kalra R, et al. Insulin Resistance and Cardiometabolic Risk Profile Among Nondiabetic American Young Adults: Insights From NHANES. J Clin Endocrinol Metab. 2022;107(1):e25–e37. DOI: 10.1210/clinem/dgab645
- Kopin L, Lowenstein C. Dyslipidemia. Ann Intern Med. 2017;167(11):ITC81-ITC96. DOI: 10.7326/AITC201712050
- Hanley JA, McNeil BJ. A method of comparing the areas under receiver operating characteristic curves derived from the same cases. Radiology. 1983;148(3):839–843. DOI: 10.1148/radiology.148.3.6878708
- Tang Y, Kuang J, Xia X, et al. Targeting microbiota-generated acetaldehyde to prevent progression of metabolic dysfunction-associated steatotic liver disease. Cell Metab. DOI: 10.1016/j.cmet.2026.01.021
- Qi X, Yun C, Sun L, et al. Gut microbiota-bile acid-interleukin-22 axis orchestrates polycystic ovary syndrome. Nat Med. 2019;25(8):1225–1233. DOI: 10.1038/s41591–019–0509–0
- Hills RD Jr, Pontefract BA, Mishcon HR, Black CA, Sutton SC, Theberge CR. Gut Microbiome: Profound Implications for Diet and Disease. Nutrients. 2019;11(7):1613. DOI: 10.3390/nu11071613
- Liu J, Wang L, Jiang M, et al. Unraveling the mechanisms of PCOS: the interplay between gut microbiota and the immune system. J Ovarian Res. DOI: 10.1186/s13048-026-02047-4
- Yang YL, Zhou WW, Wu S, et al. Intestinal Flora is a Key Factor in Insulin Resistance and Contributes to the Development of Polycystic Ovary Syndrome. Endocrinology. 2021;162(10):bqab118. DOI: 10.1210/endocr/bqab118
- Guo J, Shao J, Yang Y, et al. Gut Microbiota in Patients with Polycystic Ovary Syndrome: a Systematic Review. Reprod Sci. 2022;29(1):69–83. DOI: 10.1007/s43032–020–00430–0
- Zhou X, Li J, Guo J, et al. Gut-dependent microbial translocation induces inflammation and cardiovascular events after ST-elevation myocardial infarction. Microbiome. 2018;6(1):66. DOI: 10.1186/s40168–018–0441–4
- Heim CE, Bosch ME, Yamada KJ, et al. Lactate production by Staphylococcus aureus biofilm inhibits HDAC11 to reprogramme the host immune response during persistent infection. Nat Microbiol. 2020;5(10):1271–1284. DOI: 10.1038/s41564–020–0756–3
- Wu J, Fan C, Shi Z, Li B. The regulatory effect of Bifidobacterium dentium N8 on lipopolysaccharide-induced intestinal mucosal immune injury. Food Funct. 2026;17(4):2094–2103. DOI: 10.1039/d5fo04703j
- Rudnicka E, Suchta K, Grymowicz M, et al. Chronic Low Grade Inflammation in Pathogenesis of PCOS. Int J Mol Sci. 2021;22(7):3789. DOI: 10.3390/ijms22073789
- Harada N, Hirano I, Inui H, Yamaji R. Stereoselective effects of lactate enantiomers on the enhancement of 3T3-L1 adipocyte differentiation. Biochem Biophys Res Commun. 2018;498(1):105–110. DOI: 10.1016/j.bbrc.2018.02.198
- Lim SS, Norman RJ, Davies MJ, Moran LJ. The effect of obesity on polycystic ovary syndrome: a systematic review and meta-analysis. Obes Rev. 2013;14(2):95–109. DOI: 10.1111/j.1467–789X.2012.01053.x
- Escobar-Morreale HF. Polycystic ovary syndrome: definition, aetiology, diagnosis and treatment. Nat Rev Endocrinol. 2018;14(5):270-284. DOI: 10.1038/nrendo.2018.24
- Yu J, Ge H, Gao Z, et al. Association between body roundness index and reproductive outcomes in patients with polycystic ovary syndrome: a secondary analysis based on PCOSAct. Front Nutr. 2026;13:1705555. DOI: 10.3389/fnut.2026.1705555
- Zhao Y, Fu L, Li R, et al. Metabolic profiles characterizing different phenotypes of polycystic ovary syndrome: plasma metabolomics analysis. BMC Med. 2012;10:153. DOI: 10.1186/1741–7015–10–153
- Brinca AT, Ramalhinho AC, Sousa Â, et al. Follicular Fluid: A Powerful Tool for the Understanding and Diagnosis of Polycystic Ovary Syndrome. Biomedicines. 2022;10(6):1254. DOI: 10.3390/biomedicines10061254
